Imagine pitching a product that works, changes lives, scales globally, and still gets rejected in the same meeting for being “too small.”
That happened to Jim Fruchterman. More than once.
Jim is a Caltech-trained engineer, serial founder, MacArthur Fellow, and one of the earliest people to spot a blind spot in modern tech. There’s a massive gap between what technology can do and what venture capital will fund. And most of humanity lives in that gap.
After founding seven for-profit startups in Silicon Valley, with a very respectable failure rate, Jim realized something uncomfortable. Some of the most important problems on earth will never clear a VC investment committee, not because they’re unsolvable, but because they’re insufficiently profitable.
So he stopped trying to force them to.
Instead, he built an entirely different playbook.
The Market Failure No One Likes to Talk About
In venture land, if an idea doesn’t pencil out, it’s labeled a bad idea. End of discussion.
But Jim saw those “bad ideas” differently. They weren’t bad, they were orphaned. Markets where:
The technology already exists
Millions of people could benefit
The path to impact is clear
The revenue ceiling is real, but modest
To VCs, that’s a dead end. To Jim, it was an opportunity.
That insight led him to found Benetech, and later Tech Matters, organizations that apply Silicon Valley-grade product thinking to nonprofit markets like disability access, crisis response, mental health, education, and human rights.
The twist is not charity-first thinking. It’s business-model realism.
Nonprofit SaaS, But Built Like a Real Company
Tech Matters doesn’t build apps no one will download. Jim is allergic to that pattern.
Instead, the organization builds what is essentially vertical SaaS for underserved markets, software that organizations actually pay for, even if on a sliding scale.
One flagship example is a cloud-based platform for crisis helplines. Think Salesforce, but purpose-built for counselors supporting people in crisis. Text, WhatsApp, secure chat, integrated workflows, all designed around the realities of frontline work.
In wealthy countries, customers pay enough to generate margin. In lower-income regions, pricing drops below cost. The difference is subsidized intentionally.
The goal is not infinite growth. It’s sustainability plus reach.
A nonprofit that covers most of its budget with revenue is not a compromise. In Jim’s view, it’s leverage.
Why “Crappy Businesses” Can Be Incredible Outcomes
One of Jim’s favorite refrains is that many of his ventures would be terrible startups.
Too small. Too niche. Too slow.
And yet, they outperform the status quo by 5x or 10x in cost effectiveness. They replace obsolete systems. They unlock access. They scale to dozens of countries.
In nonprofit economics, a two or three million dollar operation that breaks even is not a failure. It’s a category leader.
This is the mental shift most technologists struggle with. Silicon Valley optimizes for outliers. Social infrastructure optimizes for coverage.
Different game. Different scoreboard.
AI Without the Hype Hangover
Jim has been working in AI since before it was cool the first time. Which is why he now spends much of his time talking people out of AI projects.
Most fail. Some fail spectacularly.
The mistake is treating AI as magic instead of machinery.
Where Jim does get excited is in boring, high-leverage applications. Automating drudgery. Summarizing case notes. Surfacing patterns. Giving frontline workers back time.
If an AI tool helps a counselor spend 10 fewer minutes on paperwork and 10 more minutes with a person in crisis, that’s real impact. Stack a few of those gains together and suddenly the same team can help 40 percent more people.
That’s not sci-fi. That’s productivity.
Open Source, But With Intent
Much of Tech Matters’ software is open source, but not for the usual reasons.
Their users don’t have engineering teams. They’re not submitting pull requests.
Open source here is about trust, resilience, and shared ownership. It’s about ensuring human rights activists know there’s no back door. It’s about disaster preparedness. It’s about making sure the software survives even if the organization doesn’t.
Revenue comes from services, hosting, and support, not licenses. The code stays open. The mission stays intact.
A Different Definition of Winning
Over his career, Jim has sold a nonprofit to private equity, incubated ventures that later became for-profit, and watched markets mature enough to sustain themselves.
He doesn’t see that as failure or mission drift. He sees it as success.
Nonprofits, in his view, can act as market-creation engines. They absorb early risk, prove demand, build infrastructure, and sometimes hand the baton to capitalism once the market is ready.
The end goal is not permanence. It’s progress.
The Bigger Pattern
If there’s a unifying lesson in Jim Fruchterman’s work, it’s this:
Technology is not inherently good or bad. Design signals intent. Business models lock in values.
Most tech history is written by companies chasing the biggest markets. But some of the most important chapters are written by people willing to build for everyone else.
The future of tech for good won’t come from softer ambition. It will come from sharper thinking, clearer economics, and the humility to admit that not every problem wants a unicorn.
Some just want to work.
👂🎧 Watch, listen, and follow on your favorite platform: https://tr.ee/S2ayrbx_fL
🙏 Join the conversation on your favorite social network: https://linktr.ee/theignitepodcast
Chapters:
00:01 — Jim Fruchterman’s Origin Story
01:05 — From Caltech to Silicon Valley Startups
02:10 — Early AI, OCR, and Reading for the Blind
03:00 — When VCs Say No to Social Impact
03:45 — The Accidental Nonprofit Insight
04:45 — Seven Startups and Choosing the Nonprofit Path
06:00 — The Market Failure Between Tech and Profit
07:10 — Applying Silicon Valley Rigor to Social Good
08:20 — Venture-Style Filtering for Nonprofit Ideas
09:30 — Distribution as the Real Bottleneck
10:30 — Introducing Tech Matters
11:15 — Nonprofit Vertical SaaS Explained
12:00 — Crisis Helplines and Cloud Infrastructure
13:30 — Competing with Salesforce in Niche Markets
15:00 — Revenue, Subsidies, and Sustainability
16:30 — Donors as Early Risk Capital
18:00 — When Nonprofits Become For-Profits
19:30 — Selling a Nonprofit and Market Creation
21:00 — Measuring Impact Beyond Vanity Metrics
22:30 — Open Source for Trust and Resilience
24:00 — What Tech Matters Is Building Next
25:30 — Mental Health Infrastructure at Scale
27:00 — AI Hype vs Real Productivity Gains
29:00 — Automating Drudgery, Not Empathy
31:00 — Technology, Ethics, and Design Intent
33:00 — Regulating Tech When It Goes Too Far
35:00 — Optimism About AI and Human Adaptation
37:00 — The Long-Term Role of Tech for Good
39:00 — Legacy and the Future of Social Impact Tech
Transcript
Brian Bell (00:01:21):
Hey everyone, welcome back to the Ignite podcast. Today, we’re thrilled to have Jim Fruchterman on the mic. Jim is a Caltech trained engineer turned serial tech for good entrepreneur who even got the MacArthur Fellowship. Wow. He founded Benetech in 1989 and more recently Tech Matters, building open source platforms for underserved communities globally. His mission, bring the benefits of technology to the 90% of humanity, typically neglected by for-profit models. Thanks for coming on, Jim.
Jim Fruchterman (00:01:46):
Glad to be here, Brian, and telling the tech for good story.
Brian Bell (00:01:48):
Yeah, I love it. Well, what is your origin story? What’s your background?
Jim Fruchterman (00:02:04):
And there was this thing called Silicon Valley going on. And everyone was going off to join a startup. And my roommate was one of the early people at Silicon Graphics. And I went, ah, heck. And so I joined a startup rocket company. It had just been legalized to have a private rocket company. And the rocket blew up in the launch pad. And I went, hmm, I think I would rather start more companies than go back to a grad program. And so I started seven for-profit companies in the Valley in a roughly 10-year period. And only five failed.
Brian Bell (00:02:32):
That’s not a bad hit, right?
Jim Fruchterman (00:02:34):
Yeah. And the two that went were both in what we now call AI, but we called machine learning or pattern recognition. And it was at the leading edge of what was AI then. And our breakthrough was to use gigantic data sets. And the application was optical character recognition, because back then grappling with language was a big deal. And the social good application was reading to the blind. Ray Kurzweil was our main commercial competitor. Ray had famously invented a reading machine for the blind 10 years earlier. His was 50 grand. Ours was five grand. And we figured, wow. So I demoed it to our venture capital board. The product worked and they went, how big’s the market? And I said, well, I think it’s, I think Kurzweil is selling about a million dollars a year. And there was this awkward silence in the boardroom. And they’re like, and the connection to the $25 million we’ve invested in this company so far, like be great PR and our employees will be really proud. And they went, nah. They vetoed it in the same meeting for excellent business reasons.
Brian Bell (00:03:29):
Yeah. I mean, is that a function of like how many people can buy it at X price basically? Right. And Kurzweil, you said it was only selling a million in units at 50,000 per unit. No, no, no, no.
Jim Fruchterman (00:03:40):
A million dollars of revenue. A revenue.
Brian Bell (00:03:42):
That’s what I mean.
Jim Fruchterman (00:03:43):
Yeah. Yeah, it was pretty weak. Now, $50,000 was a high price point. And I was wrong about the market. It actually was a $5 or $10 million market. But that wouldn’t have changed the calculus of my investors at all. Not VC backable. No. But it was really socially cool. So I went back to my lawyer. I had a software lawyer who was cheering me on. He was really, really helpful. But he heard that the board had vetoed the project. how about you start a deliberately nonprofit tech company? And I giggled because of course my then startup was accidentally nonprofit. And I went, wow, if you are a nonprofit on purpose, you’d be like successful by definition. And so he did the pro bono work. I told my wife I would do this nonprofit thing on the side for a year. And then I’d go back to my regular tech career. And that was, I don’t know, over 30 years ago.
Brian Bell (00:04:32):
Wow, so what happened next? So you did, I mean, seven, maybe you could just rattle off what those seven startups were. I’d love to hear, yeah.
Jim Fruchterman (00:04:40):
We were going to acquire one of the main math software developers. We were going to do silicon microstructures. So microfluidics on a chip, that was another one of them. Of course, I wanted to do a rocket company. No one gave us $200 million. And so, boy, now they’re sort of like in the depths of the past. And the two companies that worked were both in basically visual pattern recognition. So I started another company that did a lot of recognizing whale tails and grading coins and doing, and of course, the commercial application was inspection of products. Could you detect a defective product on a production line visually by seeing the defect? And so that company actually, I started about the same time as my nonprofit. And I was CEO of both for five years until they both grew to the point where they needed a real full-time CEO. And I chose the nonprofit because I was having a lot more fun.
Brian Bell (00:05:33):
Amazing. And so at some point, you had this aha moment. Let’s do nonprofit ventures. Tell me about that transition.
Jim Fruchterman (00:05:39):
Well, you know, the basic idea was in the Valley, if you have an idea and it doesn’t pencil out, people tell you it’s a bad idea. And I was in love with this idea, which of course, as an entrepreneur, that’s a trouble sign, but hey, I was in love with this idea. And I thought, well, what is this gap between what’s possible with technology and what’s sufficiently profitable for a venture backed or even a public company, right? There’s a bar is pretty high and it just keeps getting bigger. And so, but all the machinery of a tech enterprise is really handy. Great margins. Gosh, not a lot of the nonprofit sector has margins, right? And so, so the idea is that we could use the same power of technology, product design, user-centric approaches, all that kind of stuff and do, philanthropy five or 10 times better, just because we could see, hey, here’s this market that’s not that exciting, but invest two, four, $5 million, and suddenly I can solve this social problem for half the price or a fifth the price, or sometimes less than a 10th the price. And because the nonprofit sector is so far behind in its technology applications, because it’s relatively poor, there’s often a giant gap. Are they five, 10, or 15 years behind the times? If you’re like me and you’re paying attention to what’s going on in the Valley and in the tech industry, someone else has already figured out what that technology is good for and what it’s not good for. And I can just cherry peck and go, oh, that’s a great fit for this problem over here in the social sector. I think I will apply it. And a $2 or $3 million a year breakeven venture in the nonprofit sector, that’s a gigantic screaming success. as opposed to a dismal failure if we had had investors.
Brian Bell (00:07:17):
That’s fascinating. So how do you decide what to focus on and what kind of looks like a good problem to solve? And how do you kind of pick the tech and the problem and match that up?
Jim Fruchterman (00:07:28):
So I’d say in the early days, it was, you know, Jim comes up with a brilliant idea and then we run around and try it on people. And if they barf on it, then we don’t do it. But now it’s much more a deliberate pipeline kind of process, kind of like what a VC investor would do.
Brian Bell (00:07:43):
Or like a venture studio model almost. Yeah.
Jim Fruchterman (00:07:45):
Yeah. And so the idea is like, I have, let’s say five conversations a week with nonprofit leaders. Mostly I’m talking them out of bad ideas. I joke it’s an anti-consulting practice because, you know, you don’t need a blockchain. No one will download that app, you know, dot, dot, dot. We know we have the pattern recognition for what a bad venture looks like. We spotted a lot of them collectively if you’ve been in the industry a long time. So, but when something shows up where, You know, millions of people could benefit. The technology exists. There is a path to market. I can imagine how I get to distribution because that’s actually the Achilles heel of any venture is like, you can build it, but can you actually get it in the hands of people? If I find $5 million to invest in it, will it be generating most of its budget in revenue rather than donations in five years? Oh, and I also look for an existing reform movement or revolution. In other words, in the field, because I look at these as verticals, in this vertical, people are like, we are not doing a good enough job. And the way we did it in 1960 is not serving us so well. I want to be the tech partner to those revolutionaries and go, if you have a program innovation and you want to help millions of people, something tells me software and data might just be involved.
Brian Bell (00:08:55):
Right. And what that is called now is Tech Matters. Maybe you can tell us what that is and how you’re going about it.
Jim Fruchterman (00:09:02):
Yeah. So I led Benetech for 30 years and we did all sorts of applications in disability, education, human rights, the environment. And after leading the organization for 30 years, I kind of felt like I’d run out of ideas for disability. And so I started a new nonprofit called Tech Matters. Actually, we hosted it at my old nonprofit for the first three years. And the idea is I would have these conversations, spend a lot more time just helping the field. That’s why I wrote a book. That’s why I write a lot more these days. and mentor a lot more and get on more boards as opposed to, I have to stop doing it myself actually. But Tech Matters, the idea is that I could go into some new fields. And so pretty quickly, we found two new exciting fields within a year or two of each other. One is, it’s kind of like Salesforce for crisis response helplines. So around the world, there’s like 911 for kids in most countries, and it’s often three or four digits. And they were using on-premises PBXs five years ago. And I’m like, wait a minute, I haven’t heard that phrase like 15 years. It’s been obsolete that long. And so, you know, then you have, you’ve got telephone oriented helplines that are mired in technology that was developed last century. Let’s bring them into the cloud and let’s add text. Cause you know, young people, they text a little bit. And so, you know, add SMS, add WhatsApp, add Instagram, add secure web chat. And so, so that’s called a CELO. And so now we’re operating in 20 countries. We started in the developing world where our product kicked ass four years ago, and now we’re beating Salesforce head to head. Not because we’re as good as Salesforce, we have like 1% of the functionality, but everything we do is exactly what you would want if you’re a counseling and referral helpline dealing with people in crisis. So that’s our first big project.
Brian Bell (00:10:48):
So I might call that vertical SaaS in venture, right? You have this vertical SaaS platform, but the market’s not huge. So it’s not VC backable. It’s not gonna get to a hundred million or a billion in revenue. But you’re going to be able to get to some single digit or maybe double digit millions of revenue. And that revenue will kind of cover your costs to serve and maintain that product. That’s kind of the model. And it’s nonprofit, so you don’t have to pay tax. And everything that you do is for the benefit of the population you’re serving.
Jim Fruchterman (00:11:15):
That’s right. And we have a sliding scale pricing model. So in the US, we generate margin because, you know, we’re cheaper than Salesforce and better. And so we can actually charge people. And in Zimbabwe and Zambia, maybe we’re charging them less than our, even than our cloud costs. And we’re, I mean, we work exactly the way, if you were starting something like contact center software for dental practices or something, You would be on top of Twilio. You’d be on one of the three main cloud providers. You’d have something like Okta for single sign-on and multi-factor authentication. And so we build on top of that same stuff with people who came out of those industries, right? So they know their stuff. And we have a 10-person SaaS company that has revenue of like one and a half or 2 million, sorry, a budget of one and a half or 2 million and generates half of that in revenue right now. And we’re trying to grow that to 75% or 80%. And what we really want to do is have donors pretty much just subsidizing the developing world in a couple of years and have our core team all paid for by revenue from the US, Canada, Europe, New Zealand, places like that.
Brian Bell (00:12:23):
Interesting. So you do have donors, right? So people go, hey, I care about this problem. I’ll give you some seed capital, right? Some angel capital to kind of get started, go build this. But then it becomes self-sustaining with the revenues that you’re generating.
Jim Fruchterman (00:12:35):
That’s right. Our seed donors are not your typical nonprofit foundations because they’re pretty risk averse. It’s Eric Schmidt was our first big donor to get Tech Matters off the ground. And I get money from Jeff Skoll, first president of eBay and McGovern Foundation, which was a big tech media empire, all billionaire type fortunes, all Who get technology.
Brian Bell (00:13:00):
Tech billionaires who want to give back, right? They don’t want to just give money to VCs like me. They’re like, hey, but let’s actually fund some things that help people. I mean, technology is a force for good, I think, in the world. But I think what you identified 30 years ago is this market failure, right? Sometimes there’s a great technology that just can’t get to market to the people who really need it.
Jim Fruchterman (00:13:20):
That’s right. And there are these sustainable business models that actually work and your dollar goes a lot farther. I mean, if I’m generating half my budget in revenue, well, your dollar goes twice as far. It’s that easiest set of math. And then once the product is built by more risk-taking donors, then I can go to more traditional donors. UNICEF has become a big funder of this for obvious reasons. We’re helping lots and lots of kids. And so we had to have those early seed donors taking the risk. But once the product is starting to work, If you like what we do in Zambia, you’ll love it in Thailand, right? And so that gets to be a much easier ask. And it’s not that Jeff Skoll has stopped investing. They have two pockets, right? And I think that Jeff and Pierre Omidyar, the other main eBay guy, they kind of pioneered this idea of what’s the right capital structure for this opportunity to change the world? Sometimes it’s going to be an investment. Probably most of the time it’s going to be investment. Sometimes it might be a below market debt. thing. They’ll guarantee the first 20% of a big microcredit portfolio. So they’re going to get their money back, but maybe they’re taking more risk than they would be getting for, I don’t know, 3% interest or something. And then sometimes it’s just, nope, it’s a grant. And I am a big believer that when you sit down and think about what your value proposition is, how badly the market’s going to fail, what kind of capital you need, your product market fit, your capital stack is going to become pretty obvious, right? And if you can do it as a and make money, man, it’s a lot easier to be a profit-generating for-profit than trying to be a revenue-generating nonprofit. Because having done both, I can vouch for this.
Brian Bell (00:15:00):
Yeah. In a way, it’s almost easier to go after a big market with lots of customer pull than to try to solve all these disparate problems and Lots of different countries with technology. It’s actually a harder lift, which is why it’s not VC backable. There’s not a huge market that will have a billion of revenue.
Jim Fruchterman (00:15:17):
It’s easier to hire people. Your vendors get why they might want to partner with you. Hey, partner with me now when I’m small and someday you’ll be raking it in. I can’t make that case to Twilio.
Brian Bell (00:15:29):
Yeah. So what have been some of the challenges, other challenges with this model as you kind of, you know, try to tackle like a new problem? Is it landing the grants and the seed capital? Is it, you know, it almost sounds like you bootstrap some of these too, right? Where you’re kind of taking some of the revenue from, you know, Acelo and putting it towards the next thing.
Jim Fruchterman (00:15:52):
Yeah, a matter of fact, you know, I used to just have to do it on the side, my new venture development, right? But, you know, lots of people do that. As time has gotten on, I’ve gotten a few donors that actually kick in money so that I can actually like help people and try to figure out what my next venture is. So I’ve gotten money from, I don’t know, Okta and Splunk, you know, two tech companies that took the pledge 1% pledge back in the day, which meant that they committed 1% of their equity. They went public. Suddenly they have, you know, $10 million a year to give away. And they like what I do. So they kind of fund some of this field building work. And yeah, so I generally don’t have to raise the first 10 or 20 grand to just put someone on digging into or building a prototype and going and testing it. That’s pretty easy. But I’d say raising the capital is a challenge. It’s harder to raise nonprofit capital, especially for tech. Because your average donor wants more blankets and bags of rice, even if tech helps do it 10 times better. And then once the product is built, building the product and doing user-centered design, rapid prototyping, lean, we know how to do that. And if you’ve got product user fit, product market fit, then the challenge is, what’s the distribution play? How are we going to get this in the hands of the people that need it? knowing that we have a shoestring of marketing dollars, if that. So I’d say I’m always more concerned about distribution than, again, because I’m looking at an application that was figured out in the tech industry five or 10 years ago. And so looking at it and going, they need this or they need something very close to this, then I’m going to be more worried as, can I find the capital and can I actually come up with a distribution strategy?
Brian Bell (00:17:29):
Yeah. You know, you’re probably aware of the OpenAI saga, nonprofit, for-profit, right? And I’d love to get your comments on that. But if you’re able to and, you know, have you run into that situation yourself where like you thought it was going to be a kind of a small market and it turns out, hey, this is actually a bigger thing than we thought and we might want to flip it to for-profit?
Jim Fruchterman (00:17:46):
It did happen. And I sold one of my nonprofits to a to a private equity back roll up. And of course, the first time this guy came by, he said, Jim, I want to buy your charity, your reading machine for the blind charity. And I’m like, go away. I’m a nonprofit. And then he came back and he said, Jim, I know it’s been a couple months. Tell me your dreams. And I’m like, oh, I want to do this. I want to do that. He says, well, how about I give you $5 million and I won’t raise prices and you can go off with that $5 million and invent your next couple of projects. And so I took that deal. Yeah, I think that sometimes you can almost think about some of the nonprofit sector is like market development capital for capitalism. The ideal outcome is that kid gets educated and gets a job, right? or that person with disability.
Brian Bell (00:18:32):
Or gets fed or gets clean water, whatever it is.
Jim Fruchterman (00:18:34):
Yeah, exactly. But then starts moving up to where they become independent. They don’t need charity, right? And so the same thing goes is there’s a market failure gap here. Well, 10 years later, a roll-up that was buying the three leading companies in disability actually had a case and actually was able to raise, I don’t know, 50 million bucks, half in equity, half in debt. And I was the smallest of the three companies they bought, but it was a real deal. And that company, those products are still around 25 years after the acquisition. So it’s a small market, but they made money. And so, yeah, I do think this conversion goes on. When I start a new venture, I’m thinking, what are my exit options? And one of the exit options in almost every field is, what if it becomes attractive to be bought? And usually that’s a sign that we won.
Brian Bell (00:19:22):
Right? Yeah. So what’s, looking back, I mean, you’ve done so much. What are you most proud of?
Jim Fruchterman (00:19:28):
I think I’m proud that I broke the code on how to get technology in the hands of the 90% of humanity who need it the most and are least able to afford it, right? And it turned out I wasn’t alone, but for 10 years, I think there were a handful of us who didn’t know about each other that were all doing something similar, right? We were really excited about the product. We were making a living. We were not going to get rich. And we were all okay with that. And now there are thousands of tech teams inside nonprofits, building products, becoming the category leader because they’ve, you know, it turns on some software. You know, if you think about Kiva, people, famous microcredit, you know, It’s a software company with a sideline and micro credit. It’s like, it’s like anything in the tech industry, right? These are pull back the hood. You get a software company. This is the way the nonprofit sector is going. And that’s one of the reasons why, you know, I got motivated to write a book because there are thousands of books on how to get rich with tech. Now there has to be at least one on how to not get rich with tech, but make a big impact.
Brian Bell (00:20:26):
And what’s the name of that book?
Jim Fruchterman (00:20:27):
Technology for good.
Brian Bell (00:20:28):
Yeah. I love it.
Jim Fruchterman (00:20:29):
So the obvious title.
Brian Bell (00:20:31):
And what are some of the kind of key themes and takeaways in the book?
Jim Fruchterman (00:20:34):
Well, I have a chapter called The Bad Ideas That Everyone Tries First, the app that one will download, building a giant database in the sky because you think people are going to look at it. Well, Google beat Yahoo. Let’s not go there, right? Custom software for the million-dollar-a-year organization. That’s not going to end well, whether you’re a small business or a nonprofit. And then what are the good ideas? Design for people on the front lines, move into the cloud, which the nonprofit sector is lagging on, believe it or not. Kill the dinosaurs is one of my advice. I mean, a lot of what goes on the tech industry is either you innovate or someone’s going to innovate around you. And a lot of ventures, this is going to be the next category killer, right? We think they have a play at winning. And in tech, If you start winning, maybe you’ll get bought. And maybe that’s the way that you actually, your product gets the distribution that it really acts as a bigger player scoops you up. But in the nonprofit sector, what happens is these models calcify. I mean, my competitor, the Library of Congress was shipping audio cassette tapes through the mail less than 10 years ago. Audio cassette tapes were obsolete 30 years ago, and that was how they were doing it. And so by adopting an e-book, in this case, we adopted it, oh, I’d say 20 years ago. Early e-book, pre-Kindle, e-book, digital delivery. We were 50 more cost-effective than shipping audio cassette tapes through the mail that have been human recorded. Because we were doing e-text. And you can do a lot with e-text. So these opportunities are really exciting. And so my thing is, you should be looking for dinosaurs that are ripe to be picked off. And sometimes the dinosaur you kill might be your own. You should reinvent yourself before someone does it to you. But the nonprofit sector, the push to kill the dinosaurs is much weaker.
Brian Bell (00:22:20):
So I had in my notes that some of this has been open source. Is that right?
Jim Fruchterman (00:22:23):
Yes, a lot of the software that we do is open source, but it’s open source with a twist. Our customers are poor people and small nonprofits around the world. None of them have tech teams. None of them can contribute to the code. So the open source is more about trust and joint ownership. So, you know, you do software for human rights activists. They might want to know that it’s open source. So there’s no backdoor in there for the Russian government or something like that. And so it’s open source from a safety standpoint. We co-developed this software with the Child Helpline movement. We made it open source. It’s a disaster preparedness. What if we go away? Well, it’s open source. You hire someone who knows how to operate Twilio. There’s millions of Twilio developers out there. You’ll be able to keep it going, right? And frankly, there’s an awful lot of ways to make money in open source that aren’t about charging a license fee. And we’re disciples of that whole free movement. Find your revenue somewhere else.
Brian Bell (00:23:16):
So what are some solutions and problems you’re incubating right now that you’re excited about?
Jim Fruchterman (00:23:22):
Well, I think there’s two things that we’re working on that I’m very excited about. One is taking a salo into larger mental health. I think that right now, AI and mental health, not a great fit. But that is going to be a solved problem in the next five years. And so I believe that there’s this global unmet need for human, basically mental health services. And there’s going to be some exciting technology about there. So I think that that’s something I’m really excited about. And then the other thing that I’m also working on, because again, I keep doing too many things, is I’m starting a data governance movement. that is an alternative to the surveillance capitalism norms of the valley uh the idea is it’s called the better deal for data and it’s seven simple principles it’s kind of like the open source initiative these seven things are true it’s a better deal for data these 10 things are true it’s an open source license you know it’s like Don’t claim ownership in the data. Don’t sell it to the data brokers or give it to Meta. We’re primarily collecting it to benefit you and your community and science, but not to light our own pockets. So if we can make some money that helps us deliver the service, yay, that’s great. If we do research, we’ll strip out your PII. If it’s sensitive data, we’ll encrypt it and keep it safe. And we make a legally binding promise to these principles. And so the nonprofit sector wants to do the right thing when it comes to data. They’re not there to make a buck. And yet a lot of people, data subjects, are not trusting data collection because often they get screwed. And this started in the farming areas. who basically a big donor, research donor for ag investment, ag research, says farmers aren’t cooperating with data collection because they figure out that they get screwed. They’re going to get charged more for their inputs, paid less for their outputs, and John Deere owns the map of their farm. So if you’re going to show up, you have to have something that says you can trust us with your data. We’re here to help you primarily. So then that launches in early 2026. So it’s coming up shortly.
Brian Bell (00:25:18):
Exciting. As you look out, you know, well, maybe let’s, let’s tackle AI, right? Real quick. You know, how’s AI impacting what you’re doing right now? And how do you see it kind of impacting you in the next five years?
Jim Fruchterman (00:25:28):
Well, you know, I spend a lot of my time talking people out of AI projects, even though I love, love, love AI. I’m an AI guy from way back. And, you know, depending on who you believe, are 80% of AI projects failing or 95%. And I go, those are great numbers. The blockchain numbers are 99. One out of five are succeeding.
Brian Bell (00:25:46):
That’s great.
Jim Fruchterman (00:25:46):
Yeah, I know. Yeah. Yeah, because with blockchain and the metaverse and social good, it’s like 99.99% failure rates. I’ll take it. So I’m spending a lot of my time saying, look, don’t go around thinking that this is a magic solution. I’m a AI is normal technology guy, to quote the paper from the AI snake oil guys from earlier in 25. And so that is like, problem first, what is the pain that your user is experiencing that the AI tool is going to solve? And so, and then I spent a lot of time shining a light on the really good applications of AI in the social sector. And there’s a bunch, but they all look a little bit more boring, right? So a whole bunch of the exciting ventures I see are some variation on RAC, right? Take the off the shelf LLM, take a whole bunch of vetted content, and say, hey, LM, when you answer this question, pretend you’re the person who wrote the 5,000 answers to these 5,000 questions or base it on our, I don’t know, our journals, all the articles ever published in our journal. So I see a whole bunch of that in job coaching and helping diagnose medical issues. And then there’s, you know, your run of the mill pattern recognition. Can we we spot which of these kids is more likely to fail and get them more services? So and the thing I tell a lot of nonprofit leaders is if you see drudgery, if you see something that people are doing that they hate doing. And part of their job that they love doing, I’m guessing that the love doing part is a really high value add and the drudgery part is the low value add. Those are the things that you should be looking at AI for. And so let’s pick a crisis response. You have these counselors. The average counselor spends half their time doing data entry and half their time helping kids. If they had their druthers, they would rather be spending 80% of their time helping kids and doing 20% of data entry. oh, wow, that’s a great application of AI. What do we do? We have a summarization engine that summarizes the hour-long text chat and saves the counselor 10 minutes. They love it. And look, hey, you add two or three innovations like that, and suddenly the same staff can help 40% more kids. We’re already hitting those kinds of numbers. That’s textbook productivity gains, which is what you want to be doing with tech tools. It’s like, Invest in us and you’re going to get five or 10 X return on in your employee time and employee satisfaction is going to go up and the product, the service is going to be better. That’s what gets me excited about AI in social change.
Brian Bell (00:28:13):
Is technology inherently a good force or does it have to be kind of molded and directed towards good?
Jim Fruchterman (00:28:19):
I think design signals intent. I think that very few people start a tech company expecting to do something evil. But I have noticed a pattern. I call it, you know, committing evil by accident and then going, oh, evil pays really well. Let’s double down on evil. And I think, you know, I think that’s the Facebook story. I mean, they’d be that blunt. I mean, every time. TikTok, for that matter.
Brian Bell (00:28:41):
Right.
Jim Fruchterman (00:28:41):
Hey, the list goes on and it’s in that sort of corner. But what it really did, I’ll hate to pick on certain companies, but if you had a choice between increased child safety or make more money, I think there are certain companies that relentlessly chose make more money. So sometimes if they do enough evil, they have to actually get regulated, which is anathema in tech. But hey, if five or 10 years later, you’re as evil as alcohol or cigarettes, expect to be regulated. And I do think that we’re going to get, obviously we saw the Australian social media ban for under 16s. So society says, hey, you’ve gone too far. The tech industry is not that good at self-leasing. I’ve gone too far. But I see tons of companies that also have chosen to be really good citizens. Their product doesn’t have intrinsic social impacts that are all that negative. And the Pledge 1% companies, there’s a whole bunch of them. And I think they’re basically companies that are trying to generate value for themselves and for society and their customers individually. And I think it’s a small minority that kind of go off the rails. You know, you brought up open AI. We’re going to save the world with AI. We’re going to save the world from AI. Oh, now we’re doing erotic chatbots. That’s kind of a come down, isn’t it? Well, but there’s money there. We got a nut. Oh, well.
Brian Bell (00:29:59):
So as you look back at your career, what would you do differently if you’re starting today?
Jim Fruchterman (00:30:03):
You know, not a lot, you know? I mean, I had no idea what my career was going to end up being. And I think just being open to opportunities as they show up and what gets you so excited that you can’t not do it. So for me, becoming an entrepreneur, was the right answer for me. It’s not the right answer for a lot of people, right? I get that. But I just loved creating things, inspiring people. And I did it for a while in the tech industry. And I was proud of what I created. And then I went, hey, here’s some problems. And there’s a lot fewer competitors. Let me at that field. The nonprofit sector.
Brian Bell (00:30:44):
So thinking kind of long term with, you know, robotics and AGI and maybe even SGI coming right behind that. How do you think it’s going to impact tech and society and nonprofits and people and kind of the, you know, the poorest people that you’re trying to help? If you had a crystal ball and you’re looking out at 2035 or 2040 and we have... humanoid robots that cost 10,000 and they’re just as capable as humans. And, you know, we have like AI that’s, you know, agentically can go do any cognitive task. Like, well, how do you think that just, like, you must be pretty excited by that future.
Jim Fruchterman (00:31:17):
You know, I am an optimist overall. I do not believe that, you know, no one’s going to have a job in 2028. And I look back, no one regrets that people don’t dig... you know, sorry, don’t dig ditches. I’m going to actually start that out. Yeah, yeah, mark that. I’m going to try that again. You know, if you look back decades, no one regrets that there are no longer ditch diggers, that we use backhoes instead of pickaxes and shovels, right? No one regrets that we don’t have typing pools you know, rooms full of people doing typing or word processing. And so a lot of society’s response to technology advances are having a medium amount of time to adapt to those changes. And I actually do believe that we’re going to see the benefits of this and control the most negative aspects of this by having, I don’t know, a more abundancy sort of mindset to use a term that’s floating around a lot. in certain circles. So, you know, cheap energy, better health for people, people doing more exciting, more thought things. And I actually, I’m not a big believer that AI anytime soon is really going to understand what it’s doing. I think there’s something about wetware about biology that’s different. And I want to just make sure that we keep the technology as a tool of humans and making humans realize more of their aspirations and a better life, unless this sort of apocalyptic vision of the next species is going to take over from us. I’m a little bit less enthusiastic about that.
Brian Bell (00:32:53):
Yeah, I think we have this ability to be adaptable and to mold and direct the technology over time for good. There’ll be some creative destruction in the economy with jobs. There’ll be some negative externalities that happen, but we’ll correct, we’ll regulate, we’ll correct. You know, look at the discovery of fire. Well, that could burn down the whole village and, you know, the forest that gives us all the food and our crops. And yet it’s still very useful if we control it and understand it and mold it to our purposes.
Jim Fruchterman (00:33:23):
And I think that’s the human element. What’s our intent? What’s our goal? And let’s optimize for a better future and control the biggest downsides. Yeah, I want to go there.
Brian Bell (00:33:32):
Yeah, love it. Well, let’s wrap up with some rapid fire questions. Okay. Poker idea has most influenced you over the past year and why?
Jim Fruchterman (00:33:40):
I would say it’s the AI as normal technology article from the AI snake oil guys, a couple of AI kind of researchers at Princeton. And so they laid out a vision. And I actually believe in that as opposed to AI as magic wand.
Brian Bell (00:33:58):
Yeah. What are kind of the, what’s the thesis or the takeaway of the paper? Basically that AI is just another technology and it’s just going to be integrated into the fabric of business and society.
Jim Fruchterman (00:34:07):
Yeah. And the hard part is not inventing a new technology. This is a bigger than average sleep, but it’s just one more, you know, in a sequence. The hard part is actually getting it adopted in a way that actually generates value. And that’s what people are working on right now. We’re going to figure out those applications, but it’s not going to be this society turns over rapidly and the world’s completely different three years from now. I don’t believe that. They don’t believe it.
Brian Bell (00:34:30):
What’s one technology or capability that’s currently overhyped in social impact and one that’s underhyped?
Jim Fruchterman (00:34:35):
Kind of chatbots replacing humans. You know, I just, it’s not that good. And I mean, there was a helpline that fired all the human staff and put on a chatbot. And three days later, it was discovered saying exactly the opposite to the people who are seeking help. And they had to shut down the chatbot and they’d fired all the humans. And now there’s nothing for those people with those problems. That’s like, Ew, don’t do that. And then underhyped. I’d just say data capabilities, data savvy. So much of the nonprofit sector just doesn’t really have a handle on data. And it’s boring, but can you imagine a business person not knowing the cost to acquire a customer? or what their expected lifetime revenue is from that customer. And the nonprofit sector bumbles around without a clue. And there are actually nonprofit programs that are hurting people more than they’re helping. And the staff there don’t know that that’s true. So, oh my God, let’s measure our impact a little bit more and become a little bit more aware of what we’re actually doing that’s good, do more of that. And what we’re doing that’s bad and stop doing that.
Brian Bell (00:35:35):
So if you had unlimited resources for what project tomorrow, what would you build first?
Jim Fruchterman (00:35:39):
Ooh. Well, I think it’s the chatbot that, I don’t know, solves the 80% least severe mental health issues for 80% of the population globally. I think that would be terrific. And I can kind of see how you would get started on that path. And then the limited human resources... are deployed for the most severe cases. The person who really needs human empathy as opposed to someone to talk to because they just broke up with their significant other, boyfriend or girlfriend kind of thing.
Brian Bell (00:36:14):
What’s a misconception you regularly encounter in nonprofit tech and how do you correct it?
Jim Fruchterman (00:36:18):
That it should make enough money to be venture worthy. I can’t tell you how many tech people, you know, they sit down and they go, you’ve got, you’ve got a product, you’ve got a business. Oh, great. Let’s take it. Oh my God, this is a shitty business. I go, yeah, but it’s terrific. It’s a nonprofit because it’s going to be, you know, so much more cost effective than what we’re doing now. And they’re like, but it’s a crappy business. And I’m like, get over it.
Brian Bell (00:36:41):
Given that, what is a metric or a signal that you personally watch when you’re, when you’re building a project?
Jim Fruchterman (00:36:45):
I try to imagine if I’m the general manager or the lead product manager or whatever it is, what would I need to know to know if I was succeeding? And that’s more than just outputs. It’s more than just people we’ve touched. It’s like, what is the status quo before we got started? And how has the world changed? And I’m usually looking for a... 30% more effective, which I could translate to dollars, right? But I’m looking for five or 10x return on dollars invested, not to the investor, because we’re a nonprofit, but to society that, you know, compared to shipping more blankets or whatever the thing is, this tech investment punched way above its weight in terms of the sector. And that will often get people to switch from the old way they did things to this new way, because it was so demonstrably better.
Brian Bell (00:37:35):
So another question, follow up to that is like, how do you know when to pivot or persevere or shut down a project?
Jim Fruchterman (00:37:41):
Well, my number one thing is, you know, we built a prototype, we showed it to people and they went, Hey, this is cool, but let me tell you what my real problems are. And I’ve had projects like that. And I had to, I mean, so for every idea that actually becomes a product, I easily have 10 ideas that I thought could go. But then you actually go out and talk to people and you realize what I thought their problems were, it’s not what their problems were. That’s the number one source of this is, you know, what you think people should do with technology isn’t what they actually will do. And then there are other things. You can’t raise the money. You can’t get access to distribution. We had one project that failed because I didn’t, I underestimated political risk. We had invented, we were going to develop a humanitarian landmine detector that was much better. And what I didn’t realize is that it’s classified as an offensive munition under US export laws. And so we had to have the State Department, the Commerce Department, the Department of Defense, now the War Department, Oh, and GE, who had owned the technology, who’s licensing it to us. They all had to agree simultaneously for us to export this technology. And we could just never get them to all agree at the same moment. So we had that project. We actually got into a few hundred thousand dollars invested because we were blindsided by the fact that we needed this export license. And we just weren’t prepared and weren’t able to get it because it politically wasn’t palatable.
Brian Bell (00:38:59):
All right, last question. Looking forward to the end of your career, what do you want your legacy to be?
Jim Fruchterman (00:39:04):
I want my legacy to be people using technology throughout the nonprofit sector because we have fewer and fewer dollars in the social sector right now. And we need to do five times more. And the only way I know how to do that kind of ratio is through smart technology. I think there’s so many of these opportunities and I’m hoping to inspire a lot more people as they move through their career to maybe do a five-year stint in tech for good and get a lot out of it and do a huge amount for society.
Brian Bell (00:39:35):
Awesome. Well, thank you so much for coming on. Inspiring story and journey. I learned a lot. Thanks again, Jim.
Jim Fruchterman (00:39:42):
Glad to be here, Brian.
